Shreya Patha

About

AI Engineer with hands-on experience building Python backend applications, integrating LLM APIs, and developing RAG and multi-agent workflows. Experienced with FastAPI, REST APIs, vector search, embeddings, and structured AI pipelines using tools such as Groq, Claude, LangGraph, and FAISS.

Experience

• AIML Team Lead
  UptoSkills
  Feb 2026 – Jun 2026

Education

• Bachelor of Technology
  TKR College of Engineering and Technology
  Computer Science Engineering in Artificial Intelligence and Machine Learning · Nov 2022 – May 2026

Skills

• Deep Learning
• FastAPI
• Prompt Engineering
• Feature Engineering
• Matplotlib
• NumPy
• Pandas
• scikit-learn
• TensorFlow
• Seaborn
• JSON
• OpenAI API
• LangGraph
• Embeddings
• Semantic Search
• Claude API
• Faiss
• Groq Api

Projects

• EduBridge – AI Human-Like School Assistant
  Python, FastAPI, NLU, RBAC, Pydantic, STT/TTS, Multilingual AI

  Built a role-based AI school assistant for Students, Parents, Teachers, and School Management, supporting chat, voice interaction, and an AI avatar with lip-sync. Implemented NLU-based intent and entity extraction, strict Pydantic validation, RBAC, ownership checks, and deterministic tool routing to control access to attendance, analytics, and staff escalation features. Added support for 11 Indian languages, STT/TTS, multi-turn clarification, prompt-injection and role-spoofing protection, audit logging, and 109 automated tests.

• NAVI 360 – AI Government Notice Assistant
  React, FastAPI, Python, NVIDIA NIM, SQLite, Bhashini TTS, Docker

  Developed an AI application that processes government notices, PDFs, and images and converts complex information into plain-language explanations, deadlines, required documents, and actionable steps. Built a document analysis pipeline using vision extraction, text processing, entity and deadline extraction, AI reasoning, source verification, multilingual output, and Bhashini text-to-speech. Implemented user authentication, case management, evidence vault, deadline reminders, search, API integration, rate limiting, and Docker-based deployment.

• AI Research Assistant — Multi-Agent Research Pipeline
  FastAPI, React, LangGraph, Groq, Tavily, PostgreSQL, WebSockets

  Built a multi-agent research assistant orchestrating search, summarization, and fact-checking agents via LangGraph, integrating Groq and Tavily APIs into REST and WebSocket endpoints with live progress streaming. Designed a PostgreSQL-backed persistence layer with Pydantic-validated structured outputs for research history and report generation.

• AI-Powered Journal — Mood Tracking & Analysis App
  FastAPI, SQLAlchemy, Groq API, SQLite, Pydantic

  Built a full-stack journaling application using FastAPI and SQLAlchemy, with REST endpoints for entry creation, retrieval, editing, and deletion backed by a SQLite database. Integrated the Groq API (Llama-3) for real-time AI mood analysis, and built analytics services (weekly mood summaries, trend tracking) and data export using Pydantic schemas.

• Social Media Sentiment Analysis
  Python, Pandas

  Cleaned and processed a 20,000+ post social media dataset using Python and Pandas; performed EDA to classify positive, negative, and neutral sentiment across platforms. Built 6+ data visualizations to surface platform-wise engagement and sentiment trends.

Courses & certifications

• Workshop on Generative AI and Automation Tools · Outskill
• Business Analytics · ThinkMates
• Prompt Engineering Fundamentals
• Devengers 1.0 Hackathon